Share Email Print

Proceedings Paper

Generative synthetic adversarial network for internal bias correction and handling class imbalance problem in medical image diagnosis
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Imbalanced training data introduce important challenge into medical image analysis where a majority of the data belongs to a normal class and only few samples belong to abnormal classes. We propose to mitigate the class imbalance problem by introducing two generative adversarial network (GAN) architectures for class minority oversampling. Here, we explore balancing data distribution 1) by generating new sample from unsupervised GAN or 2) synthesize missing image modalities from semi-supervised GAN. We evaluated the effect of the synthetic unsupervised and semi-supervised GAN methods by use of 1,500 MR images for brain disease diagnosis, where the classification performance of a residual network was compared between unbalanced datasets, classic data augmentation, and the proposed new GAN-based methods.The evaluation results showed that the synthesized minority samples generated by GAN improved classification accuracy up to 18% in term of Dice score.

Paper Details

Date Published: 16 March 2020
PDF: 8 pages
Proc. SPIE 11314, Medical Imaging 2020: Computer-Aided Diagnosis, 113140E (16 March 2020);
Show Author Affiliations
Mina Rezaei, Hasso-Plattner-Institut (Germany)
Massachusetts General Hospital (United States)
Tomoki Uemura, Massachusetts General Hospital (United States)
Harvard Medical School (United States)
Janne Näppi, Massachusetts General Hospital (United States)
Harvard Medical School (United States)
Hiroyuki Yoshida, Massachusetts General Hospital (United States)
Harvard Medical School (United States)
Christoph Lippert, Hasso-Plattner-Institut (Germany)
Christoph Meinel, Hasso-Plattner-Institut (Germany)

Published in SPIE Proceedings Vol. 11314:
Medical Imaging 2020: Computer-Aided Diagnosis
Horst K. Hahn; Maciej A. Mazurowski, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?